Scientists tested a new computational method called LM-QASAS, designed to identify specific immune cells (B cells) that respond to vaccination. The method analyzes the genetic sequences of these cells to find ones that target particular vaccine components, like the spike protein of SARS-CoV-2. This study looked at data from healthy people who received mRNA COVID-19 vaccines and a separate group who received an influenza vaccine. The goal was to see if this reference-free method could accurately pinpoint vaccine-specific cells without needing external comparison data.
In the mRNA vaccine group, the method successfully identified immune cells targeting the spike protein with over 90% purity, meaning most of the cells it flagged were correct. It performed better than simpler methods based on sequence identity or abundance. The method could also accurately reconstruct how the immune response changed over time in new individuals. However, when applied to the influenza vaccine cohort, the method showed limited sensitivity, meaning it did not identify as many of the relevant cells.
The main reason for caution is that the method works best under conditions of 'robust clonal expansion,' which is a strong, focused immune response. mRNA vaccines often trigger this type of response, which may explain the high performance in that group. The weaker performance with the influenza vaccine suggests the tool's effectiveness may vary depending on the vaccine and the type of immune response it generates. This was an observational, methodological study, so it primarily shows the tool's potential in specific scenarios rather than proving broad effectiveness. Readers should understand this is a promising research tool for monitoring immunity, but its real-world application across different vaccines needs more investigation.